Chapter 7: Is growth necessary for full employment?






1st Historical introduction

2nd The production function

3rd The labour supply

4th The labour productivity

5th The labour hoarding

6th Inflexibilities

7th Environmental problems

8th Concluding remarks



1th Historical introduction


In this chapter we will critically examine the assertion that full employment can only be achieved if the growth rate of the domestic product exceeds a certain critical mark. This assertion was made primarily following certain theoretical and empirical work by Arthur M. Okun, which found its way into the economic literature as Okun's Law.


The law formulated by Okun states in fact that the unemployment rate is only reduced from a critical growth rate onwards, the so-called employment threshold. If the actual growth rate of the domestic product is below this threshold, it must even be expected that the unemployment rate will rise.


Arthur M. Okun was an American economist and Keynesian who lived from 1928 to 1980. Among other things, he studied the causes of poverty and the mechanisms that exist between domestic product and employment.


The law named after him is based on empirical studies carried out by Okun for the USA for the period 1954 to 1962, which showed the connection between the growth rate of the domestic product and the unemployment rate that is claimed in Okun's law.


In the meantime, the connections between the growth rate and the unemployment rate have been repeatedly tested empirically; these studies came to the conclusion that although certain connections between the two variables could be established, the critical threshold for the overturning of the unemployment rate was at very different values in the countries studied. (See, for example, the Wikipedia article on Okun's Law and the literature cited there).


Even though Okun's Law has entered the literature, a number of economists, such as Rudiger Dornbusch and Stanley Fischer, have considerable doubts as to whether one can really speak of a law here, since the empirical connections exhibit a high degree of uncertainty. The values of the individual parameters would fluctuate over the years and the relations were also partly unstable. Moreover, the growth of production was only determined by estimates.



2nd The production function


The relations between production and employment are generally described in the so-called production function and it is therefore natural to begin the critical analysis of Okun's Law by asking how the relations between domestic product and the unemployment rate should actually develop when taking this function as a basis.


Here, we want to base our considerations on a Cobb-Douglas production function for the domestic product, which has also been subjected to a repeated empirical test. According to this, the production function can be described as follows:



X = b * Aα * K(1-α), with



X: real domestic product b: growth factor A: labour input K: capital input

α: production elasticity related to the factor labour.


According to this, there is a certain level of real domestic product (X) at which full employment is given, i.e. a level of employment (A') is reached at which employment corresponds to a previously defined level of employment. If the domestic product is lower than this critical level in the initial period, then unemployment prevails and it requires an increase in demand and therefore also in production to reduce the unemployment rate. However, once a level of domestic product has been reached at which full employment has been achieved, then ceteris paribus full employment is likely to persist in subsequent periods if this domestic product remains constant.


Note, however, the following two limitations. We have derived this relationship under the ceteris paribus formula. This means that the domestic product depends on a large number of unspecified conditions and only if all other variables remain constant and only the level of employment varies, this established relationship between domestic product and employment will be valid. In this case, the reverse is also true, of course, that the domestic product may very well remain constant after reaching a level of full employment and that unemployment nevertheless rises.


However, then this is not already evidence for the validity of Okun's law. Nor is it proof that unemployment has risen due the growth rate of the domestic product that has remained below the critical employment threshold. It must be clarified at first on which other variables the domestic product and the employment rate depend and whether one of these variables could not simply have changed and for these reasons the unemployment rate rose. For example, during the two oil crises in the mid-1970s and the early 1980s, the price of oil rose dramatically, which in turn triggered a setback in economic activity and mass unemployment.


In the discussion of the production function, it is important to bear in mind in the second place that this is an equilibrium theory. It is not claimed that at any moment this connection formulated in the Cobb-Douglas function can be established, but only that the values of domestic product and employment on equilibrium move towards this relationship. How fast this equilibrium tendency is approached and whether this equilibrium tendency is reached in a straight line or in cyclical oscillations, is not stated in this theory at all. Therefore, it does not contradict the theory described in the Cobb-Douglas functions at all if the unemployment rate initially rises or falls under certain circumstances despite a constant domestic product.


In the following we want to deal with these 'cetera' in more detail and clarify to what extent simply a change in these additional variables has triggered the empirically established connections (in Okun's law) without actually being able to blame the missing growth rate for unemployment.



3rd The labour supply


We start our considerations with the changes in labour supply. So far, we have tacitly assumed that there is a mirror-image relationship between employment and the number of unemployed: whenever employment increases, the number of unemployed decreases by the same amount, and whenever employment decreases, the number of unemployed again increases by the same amount. However, this assumption only applies to the case where the supply of labour remains unchanged. However, we have to expect the possibility that the number of workers increases (or decreases). In this case, even if employment remains constant, the number of unemployed increases (decreases).


There are several reasons why the supply of labour may change. Firstly, due to an increase in the birth rate about 18 years ago, the number of people entering the labour force may increase more than the number of workers partially decreases at the same time due to retirement. Secondly, it is conceivable that because the average life span has risen sharply, many workers leave the labour force later. In this case, the number of unemployed would increase if the number of people entering the labour force did not decrease at the same time, or if employment increased, always assuming that the demand for labour remains unaffected by these developments.


Migration can also change the number of available workers. A reduction in the labour force potential occurs, for example, when more workers of employable age emigrate than immigrate.


If we only pay attention to the official statistics, even changed definitions of employment or unemployment can already lead to a nominal change in the reported unemployment rate. For example, it is conceivable that within the framework of the official statistics, workers who have passed the age of 60 and are considered as no longer employable are no longer listed in the statistics as unemployed, although in this example nothing has changed in the real situation of the workers concerned.


However, we must add that an increase in the domestic product caused solely by the increase in the working-age population has nothing to do with the usual concept of growth. One usually distinguishes between extensive and intensive growth. One speaks of extensive growth when the increase in domestic product was caused solely by an increase in the number of people employed; intensive growth, on the other hand, only occurs when per capita income or perhaps also labour productivity has risen. When one speaks of growth in economic theory, one generally thinks of an increase in per capita income, i.e. of intensive growth. The concept of growth is therefore used here as an indicator of an increase in welfare.



4th The labour productivity


Furthermore, the relationships between domestic product (the output of the production function) and employment (one of the input factors) only ever apply under a given technology. In the Cobb-Douglas function, these technical specifications are described by the b-parameter and the α-parameter. The b-parameter here refers to situations where all input factors are varied equally, while the α-parameter informs on the extent to which factor intensity affects the productivity of a single factor of production.


The indication that a change in labour productivity also has an influence on employment does not bring any new insights, since labour productivity is directly dependent on the technical parameters of the respective production function. Of course, it is true that an increase in labour productivity automatically leads to unemployment if the domestic product and the labour supply are kept constant. However, this statement would only provide us with new information if we could prove that productivity increases would always, or at least usually, occur without an expansion of the quantity of products; but this is precisely not the case. Technical progress does not fall like manna from heaven, but is almost always the result of investments, which in turn almost always increase production capacity.


We do distinguish between expansion investments and rationalisation investments, and in this respect the emphasis in the case of rationalisation investment is primarily on increasing productivity, whereas in the case of an expansion investment the focus is on expanding production capacity. Nevertheless, in reality almost all investments bring about some improvement in productivity, just as almost all investments are only profitable if the output is increased at the same time. In this respect, we must always reckon with the fact that productivity increases are the result of investments and that these investments usually have a positive income effect.


In general, the technical parameters in the production function depend on technical progress, and a distinction is made between neutral, labour-saving and capital-saving progress. J. R. Hicks speaks of neutral progress here when the factor intensity remains constant while the interest-wage ratio remains constant. Labour-saving progress, on the other hand, is present when the labour intensity decreases while the interest-wage ratio remains constant, i.e. there are fewer labour units per unit of capital. Finally, we speak of capital-saving progress when capital intensity decreases while the interest-wage ratio remains constant.


An increase in unemployment is therefore to be expected if there is labour-saving progress and if the savings in labour per unit of goods are not compensated by the fact that, due to the increase in investment, demand also increases and with it the supply of goods as well as the demand for labour derived from this.


Which technical progress the entrepreneurs finally choose depends in turn on the prevailing wage-interest ratio. In general, there is always a certain wage-interest ratio at which full employment can be achieved. The emergence of unemployment then depends decisively on a wrong interest-wage ratio, which impedes the full utilisation of all scarce factors of production.


In summary, one can say that if labour is saved per produced unit of goods due to labour-saving technical progress, these savings can always be compensated by a corresponding increase in the production of goods, so that, in fact, from a certain growth rate onwards, large-scale unemployment can be prevented ceteris paribus. However, it can by no means be concluded from these considerations that unemployment can be reduced and ultimately eliminated only from a very specific growth rate onwards. Pushing economic growth is at best only one way to reduce unemployment, but as we will see below, it is not even the most advisable method.



 5th The labour hoarding


Let us now turn to the question of whether Okun's Law is perhaps simply a consequence of 'labour hoarding' on the part of enterprises. By 'labour hoarding' we understand the fact that in times of economic downturn, enterprises do not lay off employees who are no longer needed for production, but continue to employ them.


Two facts are responsible for such behaviour. Firstly, it is worthwhile for enterprises not to lay off skilled workers during the recession. It is true that this results in costs that are not necessary as such, since wages and employers' social security contributions must continue to be paid, even though these workers cannot be used productively during this time.


However, if enterprises were to lay off their skilled workers at the onset of the recession, they would have to incur additional costs to recruit and train them during the next upswing, and these costs may be so high that they are still higher than the wage costs of keeping these skilled workers employed. This is especially true because in the first few years of employing a newly hired skilled worker, an enterprise incurs higher costs than the revenue gains from employing these employees. The hiring and training of a skilled worker represents an investment that only pays off for the enterprise after a few years.


Secondly, rigorous employment protection can also contribute to enterprises not laying off workers in times of recession. We speak of rigorous employment protection when these provisions do not allow dismissals even when workers are not needed at all for production due to a decline in sales. In such a case, an enterprise is faced with the alternative of not being able to dismiss workers because of the protection against dismissal, even though they are not needed for production and therefore have to incur additional wage costs, or else not being able to dismiss workers at all in times of recession. Here too, of course, almost all workers incur additional recruitment and training costs when they are hired, so that continuing to employ workers is still the more cost-effective alternative.


Now, if there is rigorous protection against dismissal, an entrepreneur will only hire new workers at the beginning of the upswing if he can count on the fact that the economic upswing has already begun. We have to be clear that from the perspective of an enterprise it is not clear at the end of a recession whether new orders are of a one-off nature or whether they initiate the reversal in the economic trend and thus the upswing.


In such a situation, an enterprise will try to carry out additional orders with overtime of the already employed workforce and may even renounce the order entirely if such a strategy is not possible.


Such a situation then leads to the economic upswing being delayed and starting later than if there were no such rigorous protection against dismissal. In any case, labour hoarding contributes to the labour market lagging behind the cyclical development of the goods markets.


One might now be tempted to see in this phenomenon a confirmation and justification of Okun's Law. In fact, in this case a small upswing, i.e. a low growth rate, is not sufficient to hire new workers and thus reduce the unemployment rate. Only when the recovery is in full swing, i.e. the growth rate has exceeded a critical level, do enterprises begin to hire new workers.


A closer look at these correlations shows, however, that Okun's Law cannot be explained by the fact of the labour hoarding alone. After all, as soon as enterprises have started hiring new workers, there is no longer any reason why the growth rate must continue in the future in order to prevent the number of unemployed from rising again. Even if the domestic product remained constant in future periods, i.e. the growth rate became zero and would in any case lay below the critical employment threshold, there would be no reason, without further additional conditions, why the achieved employment could not be maintained. According to traditional understanding, lay-offs would only be feared again if the economy collapsed anew.


Basically, rigorous dismissal legislation only has the effect of postponing the extent of unemployment. Without this protection against dismissal, entrepreneurs would have laid off more workers in total when the economy declined. Thus, even at the beginning of a new economic upswing, the unemployment rate would have been higher. Although the willingness of entrepreneurs to immediately hire more workers at the beginning of the upswing would have been greater overall without protection against dismissal, the unemployment rate would nevertheless not have been lower, since the hiring of new workers would have been based on a higher level of unemployment.


In the long run, however, rigorous protection against dismissal actually causes a reduction in the level of employment, since in the initial phase of the economic upswing entrepreneurs sometimes refrain from taking on individual orders and wait to see if there are already signs of a sustained upswing. In some cases, entrepreneurs do accept these orders, but try to fill them not by hiring new workers, but by encouraging the existing workforce to work overtime. In this case, employment increases in terms of hours worked. Nevertheless, unemployment is not reduced in this case, since the number of jobs is not increased.


6th Inflexibilities


According to the traditional understanding, macroeconomic unemployment is primarily explained by the fact that markets no longer react to changes in data - as assumed in neoclassical theory. The starting point is the theorem formulated by Jean Baptiste Say, according to which increases in the supply of goods create their own demand, so that macroeconomic unemployment cannot be explained by a lack of demand for goods, which was claimed by the proponents of the underconsumption theory (e.g. Robert Malthus). The entire sales proceeds would become income, either as remuneration for labour or for the use of other factors of production, or the remainder as profit income for the entrepreneurs. The entire income would continue to become demand; either as demand for consumer goods or it would be reserved for saving purposes, which, however, would lead to an increase in demand for capital goods. The value of total demand would therefore be exactly equal to the value of supply.


Now Say also assumed that the investment volume just corresponded to the savings sum, since it is the entrepreneurs themselves who build up savings precisely when and only when they need them for investments in their own enterprise. The danger that too much is saved does not exist, since saver and investor coincide in the person of the investing enterprise.


This assumption was certainly correct in Say's time. In the meantime, however, the income of employees also increased to such an extent that they were also able to save parts of their income. At the same time, the size of the enterprise grew with the consequence that the capital requirements could no longer be covered by the entrepreneur's own savings alone. This led to the emergence of corporations, which obtained capital on the capital market, whereby part of the savings offered on the capital market came from non-entrepreneurial households.


Although, in the course of the further development of our economies, savings and the demand for credit no longer coincided in the same person, i.e. they originated from different economic units, and although it could therefore no longer be automatically assumed that the savings sum and the investment sum were identical, the neo-classics, according to Say, still assumed that the market automatically ensured a balance between savings and investment. If a larger savings sum was offered than the demand for loans for investments, the interest rate would fall, the willingness to save would therefore decrease and at the same time the demand for loans would increase, since with lower interest rates more investment objects would become profitable.


Conversely, it applied that if the supply of savings was too low, the interest rate would automatically increase and that in this way savings would increase, but the demand for investment would decrease. Thus, if the capital markets function, the interest rate mechanism automatically ensured that the savings sum in equilibrium corresponded to the investment demand and thus that Say's theorem still applied.


This is where John Maynard Keynes' criticism of Say's theorem comes in. On the one hand, it must be assumed that not all savings are offered on the capital market, but that part of the savings is hoarded, i.e. it is not invested in a way that earns interest, and would therefore seep away. This is especially true when the economy is at its lowest point. The interest rate has fallen so much in the immediate past that interest rate increases are expected in the near future. However, interest rate increases mean price losses for fixed-interest securities. In such a situation, it becomes profitable for households not to invest their savings on the capital market. Although they then forego an interest gain, which is low anyway when interest rates are low, they also do not have to fear that they will buy securities of which the price will fall in the near future and that price losses will therefore be incurred when they sell the securities. These price losses may very well be higher than the possible interest income, so that hoarding appears to be quite profitable in these times.


On the other hand, interest rate cuts are not sufficient to induce enterprises to invest more. In times of economic downturn, production capacities are not utilised anyway and therefore enterprises have no reason to invest and thus expand production capacity. Thus, in times of recession, the interest rate mechanism fails. The market no longer ensures on its own that interest rate cuts increase the demand for investment and with it the total demand for goods and that in this way unemployment is prevented.


Now, according to the classical view, the interest rate mechanism is not the only market mechanism that ensures equilibrium on the labour markets. According to the classical view, a balance between the supply of labour and the demand for labour primarily takes place on the labour market itself. If the demand for labour is too low to employ all workers who are able and willing to work, the wage rate on a free labour market falls and with it the demand for labour rises, so that on a free and functioning labour market unemployment is reduced by itself.


Also in this issue, Keynes started from the assumption that the wage mechanism does not lead to a reduction in unemployment. Firstly, the wage rate would be predetermined anyway due to the collective agreements and therefore were fixed downwards. But even then, if we could assume downward wage flexibility as given, wage variation would have no effect on employment. The level of goods production and employment derived from it would be determined solely by the autonomous investment demand and the consumption function. The position of the consumption function, however, is not influenced by wage changes. A wage variation merely represents a movement along the given consumption function. However, an influence on production and employment can only be expected if the consumption function as such would also be shifted. This is only the case, however, if more would be consumed while income remains the same.


Keynes concluded from this analysis that full employment could only be achieved by the state having to compensate for the insufficient private demand for goods by increasing state expenditure, whereby this policy would only be successful if the additional state expenditure were financed in deficit. If the state wanted to finance the additional government spending by raising tax rates, nothing would be gained, since in this case private demand would be curbed even further and government demand would only take the place of this private demand.


Now, the discussion on Keynesian theory has shown that the market by no means fails always when it comes to reducing unemployment. It has been pointed out, for example, that besides expansion investments there are also rationalisation investments. And that enterprises will strive to improve their sales opportunities by rationalisation investments especially in times of decline in sales. The thesis that investment demand would not react to interest rate cuts in times of recession only applies to expansion investments. The demand for investment will nevertheless increase in downturn phases when interest rates are lowered, because it is precisely in these times that rationalisation is necessary.


At this point, however, we want to limit ourselves to another criticism of the Keynes doctrine. The Keynesian approach can also be criticised inasmuch as a budget deficit is by no means the only possible strategy to overcome mass unemployment. Instead of the state pursuing a deficit-oriented fiscal policy, one could also try to reduce unemployment by increasing the flexibility of the markets by political means.


It is indeed conceded here that in the current situation the markets are not in a position to bring about full employment by themselves. However, the actual cause for this deficiency is not seen in natural systemic deficits, but in the fact that this inflexibility was caused in the first place by a multitude of mainly socio-politically motivated measures.


In this way, the previous approach to avoiding mass unemployment becomes extremely questionable. In a first step, the adaptability of the markets is reduced in order to increase the social compatibility of the markets. However, one accepts here that precisely because of the reduction of the adaptability of the markets, another social policy goal is violated. If there is a lack of adaptability on the labour markets, this lack is reflected in unemployment. Therefore, further political measures are needed to reduce the unemployment that has arisen as a result of previous social policy.


But this policy also leads once again to the neglect of social goals. Thus, a state deficit policy can burden future generations if the state uses the deficit revenues for consumptive purposes.


Much would be gained already if in future, whenever social policy goals are to be realised, it was to be examined whether these goals could not also be achieved in ways without further reducing the adaptability of the markets. We can assume that in general several alternative measures are available for discussion for a political goal, which differ, among other things, in the extent to which the adaptability of the markets is impaired.


This question must also be asked for the measures that have already been implemented. We have to assume here that the adaptability of markets is determined by three determinants: Price flexibility, i.e. the question of how strong and how fast price variations become possible when imbalances occur on the markets; the elasticity of supply and demand, i.e. the question of how fast and how strong price variations trigger a change in demand or supply, whereby these reactions must proceed normally, thus e.g. a price increase must trigger a reduction in demand and an expansion of supply.


A third prerequisite is needed to avoid prolonged unemployment. We have to assume that in each period data changes occur which cause new imbalances and, on the other hand, market forces are at work which in turn reduce existing imbalances. Only when the forces that create equilibrium and the forces that reduce equilibrium are in balance, then it can be prevented that the market imbalance is not increased even further. But even here, unemployment can occur on a larger scale, namely if the market was not able to clear the markets completely in the previous period.


If this was not possible in the prior period, the current period begins with a certain magnitude of initial imbalance. Here it is not sufficient that the equilibrium-triggering and equilibrium-reducing forces are of equal magnitude; the initial imbalance can only be reduced if the imbalance-reducing forces predominate.


Now, imbalances arise through data changes. For the most part, these are entirely desirable; an increase in welfare occurs precisely by way of data changes. Either welfare is increasing because technical progress occurs, or because consumers are able to align their demand with their needs, or the data change consists of improving the economic order, e.g. by preventing disincentives or creating incentives to make production more efficient or fairer.


It is therefore certainly not desirable that data changes are prevented and that in this way the extent of data changes and thus also of imbalances is reduced. However, it is possible and desirable that the type of data changes is changed in such a way that the extent of the imbalances is reduced.


A decisive advantage of a market economy over a state planned economy is that economic decisions are made in an atomised manner, i.e. there are a large number of independent households and enterprises which react to changes in data at different times and in different ways. If, for example, the price of a security falls, the market participants may react differently to this change. Some fear that this price drop will propagate in the following periods, so they will sell their securities to avoid future losses due to further price drops. Other market participants assume that today's price drop was triggered purely by chance and that there is no real change in the economic data behind this price change, so they will even buy additional securities of this type, hoping to sell them at a price gain in future periods. Above all, the time lag that occurs from the data change to the reaction to this data change is different for different people.


Whereas a data change decreed by the state (e.g. a law) comes into effect on a very specific date and also triggers similar reactions among all affected market participants, it can be assumed in the case of a market economy regulation that some of the reactions cancel each other out, since they are directed in opposite directions and, above all, that the measures with the same direction do not occur at the same time, but occur distributed over time.


Already this stretching means that market forces are more effective. They are similar to a sewerage system that is designed to channel rainwater into the sewers in the municipalities. No matter how good a broadly designed sewer system is, it is overwhelmed in the event of a cloudburst, so that in these cases the water in the streets is not drained away. If the same amount of water had emerged distributed in time and space, the system would have been sufficient to drain the water immediately into the sewage system.


In a similar way, the existing equilibrium forces of the market would be much better able to reduce imbalances quickly if the data changes encountered a large number of decision-makers, if this caused some of the reactions to be oppositely directed, thereby causing much less imbalance, and especially if the reactions to the data change were spread out over time.


Now, it can be assumed that in a market economy there are always more independent decision-makers than in a state-planned economy. Nevertheless, these advantages of a market economy are reduced and cancelled out if only a small group of corporations have to make decisions. Here, a successful regulatory policy that prevents the emergence of monopolies can contribute to the existing and desired data changes causing fewer market imbalances or reducing them more quickly.


The legislation could also be changed in such a way that the enforcement of these laws is deliberately prolonged, i.e. that the enterprises are granted longer periods for the enforcement of these laws, so that it can be expected that the reactions and actions of the market partners will be spread out over time.



7th Environmental problems


The thesis that we can only solve the problem of unemployment with high economic growth is questionable from an ecological point of view. If it were correct, we would be condemned to permanent growth in order to guarantee full employment. However, this is exactly what is not possible at all in the very long term. We have to assume that a large part of natural resources is limited, that natural resources are consumed in production and that they cannot necessarily be replaced in every case.


This may be least true for energy resources. The traditional fossil energy resources (coal, wood, oil, gas) are limited and they are consumed in production, but new energy sources can be obtained, especially solar energy, natural gas and wind energy; in the very long term, solar energy in particular will be available for billions of years.


Nevertheless, significant bottlenecks can arise for a limited period of time if it is not possible to develop the technology for the mass exploitation of solar energy in time and to expand the long-distance power grid in such a way that solar energy and energy generated from wind power can also be brought to the places of consumption. We have to assume here that both solar energy and energy from wind power are not produced at the places of consumption, so that it will be necessary to transport electricity to the most distant places of consumption via long-distance grids.


The traditional electricity generation plants (nuclear power plants as well as coal and gas plants), on the other hand, were almost always able to produce in the areas where the electricity was also in demand. Furthermore, there is a need to expand the storage plants. While the traditional power plants were able to produce electricity evenly at all times of the year, electricity from solar and wind energy cannot be produced at all times. There are times when there is no wind and when the sky is covered in clouds and therefore, in this case, electricity production must be greatly reduced.


Furthermore, raw materials can partly be replaced by plastics and certain raw materials that are not finally consumed in production and consumption can be recovered through recycling. Overall, however, the availability of raw materials is limited and therefore limits growth opportunities.


Also, space as such is a scarce commodity and cannot be increased arbitrarily. Of course, the earth is still a long way from reaching the population density of a large city. Nevertheless, there are also spatial limits from an ecological point of view, as the sealing of the soil impedes the water balance and other natural processes.


It is a frightening thesis that full employment is supposed to be possible only with permanent growth. But also, for reasons of justice, one will hardly be able to deny today's developing countries to strive for a similarly high level of development as that achieved by today's already highly developed industrial nations. In view of the scarcity of natural resources, there is hardly any room here for sustained real growth also in the industrialised nations.


Finally, the development of demand in itself, namely the demand that does not have to be artificially awakened through advertising, should determine whether and to what extent growth is needed and desired.



8th Concluding remarks


At the beginning of this article, I pointed out that some theorists doubt whether one can actually speak of a regularity in the situations described in Okun's Law. I would like to join this criticism and also justify it a little further.


Hans Albert had criticised long ago a part of the neoclassical theorists and accused them of being addicted to model platonism. These scientists would derive hypotheses in pure models of thought by logical conclusions alone, which they already regarded as valid theories. In reality, however, these theorems did not contain any new, informative statements, they were pure empty formulas, logical derivations that resulted exclusively from the assumptions and in this respect also did not show more knowledge than was already assumed in the assumptions.


Also, these assumptions would be relegated to the data wreath, which was not to be questioned further by economists. In response to the statement of the theory of marginal productivity, that the wage rate corresponds to the marginal product of labour, which has already been refuted by empirical evidence, these neoclassical scientists would reply that they had not even claimed at all that the wage rate actually corresponded to the marginal product of labour, but rather that this theory was limited to the fact that under certain assumptions (profit maximisation, quantity adjustment and declining marginal returns) the wage rate necessarily corresponded to the marginal product of labour.


It is of course very unsatisfactory if all empirically relevant economic questions are classified as data variables and are thus basically to be answered by non-economic sciences, which in turn have set themselves completely different tasks than the problems that are to be solved within the framework of economic science. Much more efficient is a division of tasks between the individual fields of science, in which the questions raised in the context are also clarified by the part of science that investigates these questions.


One can certainly agree with this criticism; it is clear that every theoretical statement must be empirically tested before it is presented as scientific knowledge. Also, most of the models of thought developed within the framework of neoclassical theory can very well be reformulated into empirically valid, i.e. falsifiable hypotheses that then also stand up to empirical testing.


In a similar way, however, criticism can also be directed against some of the empiricists who pretend that useful theories can be developed from empirical investigations alone. Empiricism needs theory just as much as theory can only be confirmed by empirical studies.


Thus, it is quite possible that for a certain time and for a certain economic sector it is shown that there is a high correlation between two variables (e.g. growth rate and unemployment rate). Nevertheless, it cannot already be clearly concluded from these results that there is a causal connection between these two variables, i.e. that when one variable reaches a higher value, the other variable also changes in a very specific way. It would be conceivable that both variables are completely independent of each other, but that they are both directly dependent on a third variable that is not known in detail and therefore not shown, whereby this variable, for more coincidental reasons, has only occurred in this time and in this area, but will not necessarily be found at all times and in all economic sectors.


The statement of Okun's Law, according to which the unemployment rate can only be reduced if the growth rate of the domestic product is above a critical threshold, may well be true if certain other conditions are met. However, this critical growth rate is neither a necessary nor a sufficient condition for reducing the unemployment rate.


A certain growth rate is not necessary for a reduction in the unemployment rate, since even with a constant domestic product, imbalances in general and the number of unemployed in the specific sense can be reduced simply due to an increased flexibilisation of the markets.


A certain growth rate is not even a sufficient condition to reduce unemployment. It is conceivable, for example, that the economic upswing is due to labour-saving technical progress, in which the redundancy effect more than compensates for the income effect. In this case, labour-saving technical progress means that fewer workers are needed per unit of output (redundancy effect). However, since technical progress requires investment, the demand for goods and thus indirectly also the induced demand for labour is partially increased (income effect), whereby it is quite possible that this upswing is not sufficient to fully reintegrate the labour released due to rationalisation into the production process. In fact, the unemployment rate may even rise precisely due to a high growth rate.